. are all three (four?) NAPs really being used Yes. for some value of "used".
That answer is close to meaningless. Please give me your metric.
Since they are neutral exchanges, w/o policy, I can only speak for for my project. I can't speak for any of the other sites that are attached. The RA has active peering sessions at all the NAPS. We do not have active peering sessions with some of the NSP's which are receiving NSF funding through the old regionals. What is your definition of used?
. Is there any evidence that the NAPs are really backing each other up? Not sure this is possible. Perhaps the better question is, are providers using the NAPs to back each other up.
Let me rephrase. How is the NSF programmatic goal being met of creating three NAPs for redundancy purposes to avoid compartmentalization of the U.S. R&E portion of the Internet, and how is that being verified?
That question can't be answered by the NAP operators, since they don't monitor traffic. You have to ask the NSP's.
. do we have some regular examples from *any* site A initiating a connection from A to B, A to C, and A to D, where the three are verifiably (via traceroute, I guess) would traverse different NAPs (and hopefully only one each)? Yes.
So, where are they? Say, can you give me two examples for such an A-B/C/D scenario? One from SDSC, one from NSF. Your answers are a bit too flippant to me. Sorry.
Anywhere to WELL.COM (via the PB NAP) Anywhere to NAP.NET (via the AADS NAP) Anywhere to CERF.NET (via the Sprint NAP) Of course your query presumes symetric routing, which is the exception rather than the rule these days.
. Are there routing stability reports accessible online from the RA (or whoever else feels responsible for this) that graph fluctuations at the NAPs, including correlation among them? What are the quality metrics for routing stability? Being defined.
To be publicly discussed, finished, and available by ...?
Check the RPS schedule w/in IETF.
. Do all the NAPs provide online statistics? No.
Why not? Will that change? My understanding is that the NAP service providers have contractual obligations for some statistics. I know there is disagreement about what stats are appropriate, but is not there a contractual requirement for at least some baseline?
I don't think so.
. Are the NAP and RA regular reports to NSF publicly (hopefully via the Web) available? http://info.ra.net/papers have the annual report/plan papers
Are there any more reporting requirements (quarterly? Monthly?). Waiting a year per report in such a changing environment strikes me as a bit long.
There are quarterly reports, which are not yet online. It's a good suggestion and I'll investigate.
If I wanted a comprehensive snapshot of the current state of the NAP-union, where/how would I get it.
Not enough info. You can dump the in-addr zones to discover who has been assigned an IP address You can see who has signed an MPLA, if they exist. You can check the CERFnet MAP You can check out http://info.ra.net/div7/ra/ep.html All of these have different viewpoints on the "NAP-union" and none have what I think you want. --bill